COOKED
The clerical core of this job — checking materials in and out, issuing library cards, sending overdue notices, entering catalog records, tracking holds and interlibrary loan paperwork — has been eroding for two decades via self-checkout, RFID sorters, and integrated library systems, and AI closes the remaining gap on reference lookups and record cleanup. What holds is the physical and human side: shelving and shelf-reading, running story hours and after-school desks, helping patrons who cannot navigate a screen, and being the person in the building when the printer jams or a patron needs help. There is no license and no personal liability, and budget-driven staffing cuts, not technology alone, will set the pace of decline.
Roughly flat across the period, with year-to-year wobble.
Median pay $27,490 → $36,910 +7.4% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
-6.7% 84,500 → 78,900 on the projections basis
Exposed, and shrinking
Both signals point the same way: the tasks are largely automatable and the BLS projects -6.7% by 2034. This is the case where the score and the forecast agree, and it is the one worth taking seriously.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~12,800 openings a year on average, including replacing people who leave.
PageBook SorterLibrary AideLibrary PageLibrary ClerkMedia AssistantMicrofilm ClerkStacks AssistantBookmobile DriverCirculation ClerkLibrary AssistantLibrary AssociateFilm Library ClerkLibrary SpecialistRegistration ClerkShelving AssistantLibrarian AssistantReference AssistantTechnical AssistantCataloging AssistantSubstitute LibrarianCirculation AssistantAcquisitions AssistantLibrary Media Assistant
Holding it up: embodiment . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier A 7 reflects that maybe a third of the day — shelving carts by call number, shelf-reading for misfiled items, setting up the story-hour room, unjamming the public printer, walking a patron through the OPAC or a resume template — cannot be scripted away, while circulation desk transactions, overdue notice generation, MARC record copy-cataloging, and ILL request routing already run themselves in Koha/Sierra/Polaris with self-check kiosks and RFID sorters absorbing the volume.
Some physical or field component At 10, the physical work is real but confined to a climate-controlled building on known floorplans: lifting and pushing loaded book trucks, reaching high and low stacks, processing and repairing damaged spines, hauling donation boxes and AV equipment for programs — repetitive and bodily, but nothing like a lineworker or field tech operating in weather and unmapped conditions.
No licence, no signature requirement A 0 is literal: the MLS-holding librarian handles collection decisions and the challenged-material process, most assistant postings ask only for a high school diploma with on-the-job training, and no state licensing board, certification, or personal exposure attaches to checking out a book or entering a bib record.
Executes defined procedures on defined inputs A 4 fits work governed by the circulation policy manual and the fine schedule: fee waivers, hold placement, meeting room bookings, and lost-item charges all have written thresholds you apply and escalate above, and the genuinely ambiguous calls — a records request, a patron behavior incident, a materials challenge — go straight to the librarian or branch manager.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (7/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (0/20) is whether the law requires a licensed human to sign. Trust premium (8/20) is whether buyers specifically pay for a person. Judgment and accountability (4/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 12 of this occupation's 29 points (41%).
Embodiment (10/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
Self-Enrichment Teachers EXPOSED
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 42/100 — EXPOSED.
Continued shift of the role toward physically-present duties automation can't touch: shelving and shelf-reading in dense open stacks, processing donations, setting up and running story hours and maker/3D-print spaces, unjamming printers, wrangling public-computer users. If systems automate the desk transactions entirely, the residual job is the in-building tier; watch for job postings that drop 'circulation' language and lead with programming and stacks work.
Digital-navigator and benefits-access funding that pays specifically for a human sitting beside a patron: IMLS/state-library digital equity grants, and the pattern of public libraries being contracted as in-person assistance points for government services (e.g. IRS/VITA tax help, ACA and SNAP application help, passport acceptance agents). If a library assistant is the designated human at such a counter, buyers are paying for presence, not lookup.
De facto frontline duties that carry real consequence: mandated-reporter obligations for unattended minors, deciding whether to call police or a social worker on an intoxicated or in-crisis patron, enforcing behavior policy and issuing suspensions. Formalizing these in job descriptions and training (as several urban systems have done with social-work partnerships and trauma-informed training) moves the role from clerk to on-scene decision-maker.
Two-tier split is real but thin: if AI absorbs record cleanup, overdue notices, and simple reference, what remains is programming design, patron de-escalation, and troubleshooting — but this remainder is small enough that libraries can cut headcount rather than redefine the role. Rises only where a system commits to keeping the position and rewriting it around programming.
The limit. No licensure, no personal liability, and no credible route to one — omit liability_shield entirely. The binding constraint is municipal and school-district budgets, not capability: a redefined, judgment-heavier role is likelier to be filled by a lower-headcount librarian (25-4022) or a part-time page than by a retained clerical assistant. Realistic ceiling is low-40s and even that requires deliberate institutional choice.
| New York-Newark-Jersey City, NY-NJ | 7,080 | $37,860 +3% |
| Chicago-Naperville-Elgin, IL-IN | 4,420 | $35,160 -5% |
| Los Angeles-Long Beach-Anaheim, CA | 3,840 | $47,770 +29% |
| Boston-Cambridge-Newton, MA-NH | 2,130 | $45,890 +24% |
| San Francisco-Oakland-Fremont, CA | 1,720 | $63,000 +71% |
| Philadelphia-Camden-Wilmington, PA-NJ-DE-MD | 1,690 | $37,350 +1% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 1,290 | $46,820 +27% |
| Cincinnati, OH-KY-IN | 1,200 | $37,200 +1% |
| San Francisco-Oakland-Fremont, CA | 1,720 | $63,000 +71% |
| San Jose-Sunnyvale-Santa Clara, CA | 840 | $62,780 +70% |
| Santa Maria-Santa Barbara, CA | 100 | $54,590 +48% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 29. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.